Use expontential backoff algo for internode reconnections (minio#17052)

hushuitian · May 2, 2023 · 4640b13 · 4640b13
1 parent 1704aba
commit 4640b13
Showing 1 changed file with 42 additions and 13 deletions.
diff --git a/internal/rest/client.go b/internal/rest/client.go
@@ -75,9 +75,9 @@ type Client struct {
 	// is online or offline.
 	HealthCheckFn func() bool
 
-	// HealthCheckInterval will be the duration between re-connection attempts
-	// when a call has failed with a network error.
-	HealthCheckInterval time.Duration
+	// HealthCheckRetryUnit will be used to calculate the exponential
+	// backoff when trying to reconnect to an offline node
+	HealthCheckReconnectUnit time.Duration
 
 	// HealthCheckTimeout determines timeout for each call.
 	HealthCheckTimeout time.Duration
@@ -309,14 +309,14 @@ func NewClient(url *url.URL, tr http.RoundTripper, newAuthToken func(aud string)
 	// Transport is exactly same as Go default in https://golang.org/pkg/net/http/#RoundTripper
 	// except custom DialContext and TLSClientConfig.
 	return &Client{
-		httpClient:          &http.Client{Transport: tr},
-		url:                 url,
-		newAuthToken:        newAuthToken,
-		connected:           online,
-		lastConn:            time.Now().UnixNano(),
-		MaxErrResponseSize:  4096,
-		HealthCheckInterval: 200 * time.Millisecond,
-		HealthCheckTimeout:  time.Second,
+		httpClient:               &http.Client{Transport: tr},
+		url:                      url,
+		newAuthToken:             newAuthToken,
+		connected:                online,
+		lastConn:                 time.Now().UnixNano(),
+		MaxErrResponseSize:       4096,
+		HealthCheckReconnectUnit: 200 * time.Millisecond,
+		HealthCheckTimeout:       time.Second,
 	}
 }
 
@@ -337,6 +337,28 @@ func (c *Client) LastError() error {
 	return c.lastErr
 }
 
+// computes the exponential backoff duration according to
+// https://www.awsarchitectureblog.com/2015/03/backoff.html
+func exponentialBackoffWait(r *rand.Rand, unit, cap time.Duration) func(uint) time.Duration {
+	if unit > time.Hour {
+		// Protect against integer overflow
+		panic("unit cannot exceed one hour")
+	}
+	return func(attempt uint) time.Duration {
+		if attempt > 16 {
+			// Protect against integer overflow
+			attempt = 16
+		}
+		// sleep = random_between(unit, min(cap, base * 2 ** attempt))
+		sleep := unit * time.Duration(1<<attempt)
+		if sleep > cap {
+			sleep = cap
+		}
+		sleep -= time.Duration(r.Float64() * float64(sleep-unit))
+		return sleep
+	}
+}
+
 // MarkOffline - will mark a client as being offline and spawns
 // a goroutine that will attempt to reconnect if HealthCheckFn is set.
 // returns true if the node changed state from online to offline
@@ -347,8 +369,14 @@ func (c *Client) MarkOffline(err error) bool {
 	// Start goroutine that will attempt to reconnect.
 	// If server is already trying to reconnect this will have no effect.
 	if c.HealthCheckFn != nil && atomic.CompareAndSwapInt32(&c.connected, online, offline) {
-		r := rand.New(rand.NewSource(time.Now().UnixNano()))
 		go func() {
+			backOff := exponentialBackoffWait(
+				rand.New(rand.NewSource(time.Now().UnixNano())),
+				200*time.Millisecond,
+				30*time.Second,
+			)
+
+			attempt := uint(0)
 			for {
 				if atomic.LoadInt32(&c.connected) == closed {
 					return
@@ -362,7 +390,8 @@ func (c *Client) MarkOffline(err error) bool {
 					}
 					return
 				}
-				time.Sleep(time.Duration(r.Float64() * float64(c.HealthCheckInterval)))
+				attempt++
+				time.Sleep(backOff(attempt))
 			}
 		}()
 		return true